Summary
- Answers 33 of 80 verification questions correctly (41.2%), placing #54 of 72 models tested on the current suite.
- Generates 107 output tokens per second, faster than 55% of the models we measure.
- Starts responding after 0.7 s (median time to first token), including thinking time.
- Costs $0.00053 per question as billed, #9 cheapest of 12 models priced under $0.15 per 1M tokens.